In article .com,
says...
In our robot drivetrain we have a set of 20 DP (3/8" face width) gears
with a 2-speed shifting mechanism. The gears (high grade steel, not
hardened) are bored and reamed to .750" and have 4 semicircle cutouts
90 degrees apart. Each semicircle is approximately .125" deep. The
gears ride on a stainless steel shaft with 2 sets of holes 90 degrees
apart around the shaft and the center of the shaft bored out
approximately 1/2". Ball bearings ride in each hole on the shaft and a
plunger is pulled back and forth inside the shaft. To engage a
particular speed, the plunger is pulled into alignment with one set of
holes and ball bearings (1/4" dia stainless steel, precision ground)
are pushed out to mesh with the semicircles cut into the bore of the
gear. The other gear that is not engaged free rotates on the shaft.

The problem we have been having is that on one of the gearboxes
something has caused wear on the shaft which leads to the non-engaged
gear unable to freely rotate. This leads to the equivalent of both
speeds being engaged at the same time (not good) and effectively stalls
the motor. We have taken apart the gearbox and replaced the shaft,
gears, and bearings with spares, and made sure that both the gears and
the shaft are clean and lightly lubricated. The gearbox is enclosed so
there is no foreign material that could be getting into the gearboxes.
The wear seems to be build up of metal from the gears onto the shaft.
This is not a flaw in gearbox design, as the system worked for many
hours of competition and testing earlier in the season. In addition,
one gearbox has never had this problem, and the other gearbox has done
this twice in two days.

What you're seeing is galling, not wear, and is a materials rather than
a design problem. The best solution would be to replace one of the
sliding components with a better bearing material. If the shaft is a
diificult part to make, I'd consider a bronze insert in the bore of the
gear. Regular SAE 660 bearing bronze is easy to get hold of and easy to
machine if the loads are low. One step up would be a phosphor bronze.
The proper aluminum or manganese bronze would be at least as strong as
the gear.

Enter "bronze" in the search window at http://www.mcmaster.com/ and
you'll get an overview of the common bronzes. More detailed info is
available at Matweb.

http://www.matweb.com

Ned Simmons